Altered vertebrae morphology and bone mineralization in a ...Jul 10, 2020 · 3 56 Introduction 57...
Transcript of Altered vertebrae morphology and bone mineralization in a ...Jul 10, 2020 · 3 56 Introduction 57...
1
Altered vertebrae morphology and bone mineralization in a zebrafish model of CHARGE 1
syndrome 2
Maximilian Breuer1*; Maximilian Rummler2*; Charlotte Zaouter1; Bettina M. Willie2#; 3 Shunmoogum A. Patten1,3# 4
*, # authors contributed equally 5
Affiliations: 6
1. INRS – Centre Armand Frappier Santé Biotechnologie, 531 Boulevard des Prairies, Laval, 7 QC, Canada, H7V 1B7 8 2. Research Centre, Shriners Hospital for Children-Canada, Department of Pediatric Surgery, 9
McGill University, 1003 Decarie Blvd, Montreal, Canada H4A 0A9 10
3. Centre d'Excellence en Recherche sur les Maladies Orphelines - Fondation Courtois 11
(CERMO-FC), Université du Québec à Montréal (UQAM), Montréal, QC , Canada 12
Key words: CHARGE syndrome, microCT, Zebrafish, late onset idiopathic scoliosis, 13 osteogenesis 14
*Correspondence: Correspondence should be addressed to 15 Dr. Maximilian Breuer 16 INRS- Centre Armand-Frappier Santé et Biotechnologie 17
531 Boulevard des Prairies 18 Laval, Quebec 19
H7V 1B7 20 Canada 21
[email protected] 22 +1 (438) 341-0445 23 24
Conflict of Interest 25
The authors declare no conflicts of interest 26
27
28
29
30
31
32
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
2
Abstract 33
CHARGE syndrome patients present features of idiopathic scoliosis in over 60% of cases, 34
reduced bone mineral density and in a few cases osteopenia. While several clinical cases and 35
studies regarding the spinal deformities in CHARGE syndrome bearing CHD7 mutations are 36
well-documented, the underlying mechanisms remain elusive. Here, we detect and 37
quantitatively analyze skeletal abnormalities in adult and young chd7-/- larvae. 38
We show that young chd7-/- larvae present with scoliosis and kyphosis already at 9 dpf. Gene 39
expression analysis confirmed the reduction of osteoblast markers and Pparγ targets. MicroCT 40
analyses identified abnormal Weberian apparatus structure and vertebral body morphology in 41
chd7-/- mutants, with highly mineralized inclusions, along with variances in bone mineral 42
density and bone volume. Additionally, we detect a specific depletion of Col2a1a in the 43
zebrafish vertebral cartilage, in line with a significantly reduced number of chondrocytes. 44
Our study is the first to elucidate the mechanisms underlying morphological changes in 45
vertebrae of adult chd7-/- zebrafish and decreased spinal integrity. The chd7-/- zebrafish will be 46
beneficial in future investigations of the underlying pathways of spinal deformities in CHARGE 47
syndrome. 48
49
50
51
52
53
54
55
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
3
Introduction 56
Impaired bone development and spinal dysmorphisms are a major health concern in various 57
genetic disorders. Among these, Idiopathic scoliosis is a commonly observed phenotype and 58
the most common skeletal abnormality in children.(1) Idiopathic scoliosis is a curvature of the 59
spine with unidentified genetic cause. However, many risk genes have been connected to the 60
underlying mechanisms.(2) One of these related genetic disorders is the congenital 61
multisystemic CHARGE syndrome (CS) named after the major characteristics of Coloboma, 62
Heart defects, Atresea chonae, Retarded growth, Genital and Ear abnormalities.(3-5) Though, 63
considering the wide variety of observed phenotypes, the list of characteristics has been 64
continuously augmented to include craniofacial abnormalities and spinal deformations.(6-8) 65
Along these lines, CS is closely associated with skeletal deformities such as idiopathic scoliosis, 66
kyphosis and hemivertebrae.(9) In fact, idiopathic scoliosis is observed in a majority of CS 67
cases, with studies showing over 60% of patients having diagnosed scoliosis at an average age 68
of just over 6 years.(10-12) In some cases, the areal bone mineral density (aBMD) is reduced. (13) 69
CS is most commonly caused by a mutation in the chromodomain ATP-dependant helicase 7 70
(CHD7). Mutations are distributed evenly throughout the gene and the vast majority of cases 71
are sporadic, with only very few cases of familial CS.(7, 14-16) Analysis of pathways related to 72
CHD7 show involvement in neural crest differentiation/proliferation/migration and stem cell 73
quiescence.(17-19) Furthermore, studies focusing on CS have identified regulatory mechanisms 74
of CHD7 in the immune response and more strikingly in brain development.(20, 21) Most 75
recently, CHD7 has been connected to abnormal GABAergic development resulting in an 76
autistic-like behavior (Jamadagni et al., unpublished data). While the function of CHD7 has 77
been extensively studied, little attention has been given to the underlying mechanisms in 78
skeletal development.(22) Some studies have directly linked CHD7 to osteogenesis. (23, 24) 79
Investigation in cell cultures have shown a dependency on CHD7 for successful differentiation, 80
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
4
more specifically to a differentiation complex with SETD1B, NLK and SMAD1 resulting in 81
depleted osteogenesis upon depletion of CHD7 in favour of adipogenesis by regulating PPARγ 82
target genes.(23, 24) Various models to investigate this role in vivo have been proposed. A mouse 83
model for Chd7 deficiency, termed “looper” presents with ear ossicle malformations, but no 84
spinal deformities.(25) However, mouse models for Chd7 deficiency have limitations, as null 85
mutants are embryonically lethal by 10 days.(26) Yet, a recent zebrafish model has proven 86
valuable in the modelling of chd7 dependent CS revealing a reduction in vertebrae 87
mineralization of young larvae.(27) 88
Zebrafish have become increasingly relevant in the study of fundamental bone development 89
and bone related disorders(28-30), including scoliosis, osteoporosis, age related osteoarthritis.(31, 90
32) Notably, the spinal structure in zebrafish has been characterized in detail and comprises three 91
major regions: the Weberian apparatus, a specialized structure consisting of the first four 92
vertebrae connecting the auditory system and the swim bladder to amplify sound vibrations, the 93
precaudal vertebrae which are connected to neural arch and spines to form the chest and, the 94
caudal vertebrae of the tail region with neural and hemal arch.(33) Simplicity of analysis in 95
zebrafish to investigate spinal structures has been used to understand the effects of mechanical 96
loading on bone mineralization.(30, 34) Zebrafish have specific advantages in the analysis of 97
skeletal development such as the closely related structure to humans, rapid bone development 98
with first mineralization of vertebrate occurring after only 5 days post fertilization (dpf), as well 99
as ex utero development and transparency of young larvae simplifying the use of in vivo staining 100
and transgenic lines to allow for the analysis of early calcification in zebrafish bony 101
structures.(35-38) Furthermore, the simplicity of high-throughput drug screening in this teleost 102
model makes the model highly useful for investigating new drug targets and effect on skeletal 103
development.(39, 40) These advantages have raised interest due to their high remodelling 104
efficiency of skeletal structures including their possibility to regenerate fins, as well as the 105
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
5
effectiveness to reproduce spinal deformity phenotypes.(41, 42) Some of these studies in zebrafish 106
show a close link to the Collagen Family.(43, 44) These highly conserved pathways between 107
zebrafish and humans allow us to investigate the underlying pathomechanisms linking CHD7 108
to idiopathic scoliosis and other skeletal abnormalities. 109
We recently generated a stable chd7-/- zebrafish mutant using CRISPR/Cas9 that replicates 110
hallmarks of CS (Jamadagni et al., unpublished data), consistent with our previous findings 111
using a transient chd7 morpholino knockdown model.(27) In this study we describe for the first 112
time a detailed analysis of skeletal development in a zebrafish chd7-/- mutant model for CS. 113
Early larvae screens reveal a delay of mineralization of vertebrae bodies linked to deficient 114
osteoblast differentiation. Gene expression and IHC analysis reveals striking reduction in 115
osteoblast, chondrocyte, and collagen matrix markers, particularly displaying diminished levels 116
of Col2a1. Finally, extensive MicroCT analysis in adults show abnormal mineralization and 117
morphological structures in precaudal and caudal vertebrae. 118
Results 119
Zebrafish chd7-/- larvae show spinal deformities 120
To investigate if spinal deformities are present at an early developmental stage, we screened 121
young chd7-/- larvae at 9 dpf (n=110) for morphological abnormalities. 39% of chd7-/- larvae 122
exhibited highly varying, scoliosis-like phenotypes at both precaudal and caudal regions of the 123
spine (Fig. 1a). Additionally, some of these also presented with a kyphosis-like phenotype 124
(Supplemental Fig. 1a). Even with the morphological phenotype, fish were reactive to a touch 125
response. 126
chd7-/- larvae have a spinal mineralization deficit that is nutrient dependent 127
Given the spinal deformities in 9dpf chd7-/- larvae, we next sought to investigate the efficiency 128
of mineralization during larval development. Calcein staining for fluorochrome labeling of 129
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
6
bones revealed no delay in the onset of mineralization of the first precaudal vertebrae at 6 dpf. 130
Upon further development, at 9dpf, chd7-/- larvae showed a significant reduction in the number 131
of calcified vertebrae (Fig. 1c). Notably, chd7-/- larvae were less efficient in complete 132
calcification of vertebrae towards the posterior region of the spine (Fig. 1b, c). 133
Since calcification of the bone matrix is dependent on mineral uptake from food, we tested how 134
the onset and development of vertebral bone structure calcification in 9 dpf chd7-/- zebrafish 135
was dependent on nutrition accessibility. Notably, wildtype and chd7-/- larvae showed no 136
difference in the number of mineralized vertebrae in a starvation situation between 5 dpf and 9 137
dpf. Expectedly, we observed significant differences when food was available from 5 dpf until 138
9 dpf, with chd7-/- zebrafish having a reduced number of mineralized vertebrae compared to 139
controls. The chd7-/- larvae had the same number of mineralized vertebrae when food was 140
available or absent (Fig. 1b, d). Starvation in zebrafish larvae between 5 dpf and 9 dpf also 141
resulted in smaller yolks, which is to be expected, considering the high dependency on nutrition 142
supply by the yolk. Surprisingly, we found that Weberian vertebrae, which are the first 4 143
vertebrae of the spinal column, were more rapidly calcified in chd7-/- larvae than in controls 144
(Fig. 1b arrows, Fig. 1e). 145
We investigated if mineralization was affected in later development. Insufficient mineralization 146
was also evident in 4-week-old juvenile zebrafish with spinal deformities. Alizarin red staining 147
revealed decreased mineralization of caudal vertebrae, with striking deficiency toward the 148
vertebral body (Fig. 1f). 149
Osteoblast differentiation is impaired in chd7-/- larvae 150
CHD7 has been shown to regulate osteogenesis by controlling PPARγ to promote skeletal 151
development and regulate expression of promoting genes. To understand if the Pparγ pathway 152
is directly involved in the observed bone mineralization deficiency and abnormal morphology 153
in chd7-/- larvae, we tested for Pparγ target gene expression and osteoblast differentiation 154
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
7
marker genes at the onset of the defect at 9 dpf (Fig 1g). The key target for Pparγ, runx2a and 155
runx2b, were significantly downregulated in chd7-/- larvae compared to controls. Further 156
markers for osteogenesis were affected, including a significant downregulation of osteoblast 157
markers sp7/osterix, bglap and postna. The chd7-/- larvae also had significantly lower levels of 158
osteocyte markers and other genes regulating mineralization such as acp5a and sost, while mgp 159
remained unchanged compared to control larvae. Additionally, we detected a significant 160
upregulation of osteoclast marker ctsk in chd7-/- larvae. 161
Abnormal development of the Weberian apparatus in adult chd7-/- zebrafish 162
To assess the effects of chd7 deficiency on vertebral growth and maturation, we analyzed 163
vertebrae morphology and mass using microCT in adult zebrafish. Morphologically, adult chd7-164
/- zebrafish were unchanged in body length, however, appeared strikingly leaner in body stature 165
compared to adult control zebrafish. 166
Since mineralization in the young larvae varied from anterior to posterior, we decided to 167
analyze the three major structures of the zebrafish spinal column: The Weberian apparatus, 168
precaudal and caudal vertebrae. The Weberian apparatus and its specialized structures, the 169
supraneurals, intercalarium, tripus and parapophysis, were thinner and smaller in morphology 170
in chd7-/- adults in comparison to controls (Fig. 2 b). We further performed a detailed analysis 171
of volumetric bone mineral density (vBMD), bone volume (BV), total volume (TV), and bone 172
volume fraction (BV/TV) of 1-year-old chd7-/- larvae and control zebrafish. We did not detect 173
any significant differences in vBMD, BV or TV (Fig. 2 c-h). However, we found a significantly 174
greater variance in TV of both intercalarium and parapophysis, as well as in BV of the 175
parapophysis in 1-year-old chd7-/- zebrafish, as determined by F-test analysis (Supplemental 176
Table 1). The same results were observed in 2-year-old chd7-/- compared to control zebrafish 177
(Supplemental Table 2). 178
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
8
Abnormal mineralization and malformations in the body end plates of precaudal vertebrae in 179
adult zebrafish 180
Vertebrae structures were analyzed by assessing key angles of the structure (Fig 3a) and 181
mineralization features. MicroCT analysis of precaudal vertebrae identified abnormal patterns 182
of mineralization and abnormal structures, that were especially pronounced at the arches and 183
transverse processes in 1-year old chd7-/- and control zebrafish (Fig. 3b, c). Precaudal vertebrae 184
showed significantly higher overall vBMD, while the distribution of mineralization between 185
the body and arch remained unaffected (Fig. 3d, e). However, structurally, precaudal vertebrae 186
of chd7-/- mutants show an increase of bone volume towards the body end plates of vertebrae, 187
which in most cases show major malformations manifesting as highly mineralized inclusions 188
(Fig. 3b, arrows). Additionally, vertebrae in 1-year-old chd7-/- zebrafish presented with a larger 189
BV/TV compared to controls (Fig. 3g-i). Analysis of the vertebrae neural arch revealed 190
distortion with a significantly larger rising angle, as well as a larger variance of the angles (Fig. 191
3f). Vertebrae showed a significantly changed body angle and significant variance of body 192
angle and vertebrae length (Supplemental Table 3). While a depletion of vBMD in the arches 193
and abnormal vertebrae body structure was observed in 2-year-old fish, the effects seen on total 194
vBMD and BV were not apparent in the few samples that survived until this stage 195
(Supplemental Fig. 2 and Supplemental Table 4). Notably, one of our chd7-/- mutants showed 196
vertebral fusion of the analyzed precaudal vertebra and following vertebra, which was not 197
observed in any of the controls (Supplemental Fig. 1). This further demonstrates the wide 198
variation in severity of the bone phenotype in chd7-/- zebrafish. 199
Abnormal caudal vertebrae in chd7-/- zebrafish vary in arch angles and reduced mineralization 200
We next determined key structural features (Fig. 4a) and mineralization characteristics of 201
caudal vertebrae. The chd7-/- zebrafish presented with warped hemal and neural arches and 202
abnormal body structure compared to controls (Fig. 4b). Similar to precaudal vertebrae, the 203
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
9
growth zones of the body end plates were enlarged and showed inclusions of highly mineralized 204
matrix (Fig. 4b). Unlike control zebrafish, chd7-/- mutants had abnormal vertebrae body 205
structure, which usually exhibited an hourglass shape and resulting in greater variance of 206
measured body angles (Supplemental Table 3). 207
In line with the neural arch screened in the precaudal vertebrae, the hemal arch in the caudal 208
vertebrae showed significantly wider arch rising angles (Fig. 4f), while the neural arch 209
displayed significantly greater variance (Fig. 4g) in 1-year-old chd7-/- zebrafish. In contrast to 210
precaudal vertebrae data, the caudal vertebrae showed significantly larger variance of Total 211
vBMD and of both arches and vertebral body vBMD (Supplemental Table 3). A significant 212
insufficiency in mineralization of the arches compared to the vertebrae body was also measured 213
in chd7-/- zebrafish compared to controls (Fig 4 d, e). Further analysis also revealed a significant 214
variance in BV, TV and BV/TV in chd7-/- zebrafish compared to controls (Fig h, i). 215
While significantly distorted arches were still observed in 2-year old mutants, the effects on 216
vBMD and BV were not noticed in the few surviving fish that could be screened (Supplemental 217
Fig. 3 and Supplemental Table 4). 218
Col2a1 of the ECM is depleted in chd7-/- zebrafish 219
To obtain a more complete analysis of vertebrae integrity, we analyzed the extracellular matrix 220
(ECM) structures of the vertebral column in 9 dpf larvae and 1-year-old chd7-/- zebrafish and 221
controls. Hematoxylin-Eosin (HE) staining of precaudal vertebrae sections showed a significant 222
reduction in the number of nuclei measured in the vertebral cartilage of 1-year-old chd7-/- 223
zebrafish and controls (Fig. 5a, b). More detailed analysis using Safranin/Fast Green O staining, 224
revealed a significant reduction in the number of chondrocytes in the vertebral cartilage in chd7-225
/- zebrafish in line with the HE staining (Fig. 5c, d). Additionally, both stainings indicate a 226
hypertrophic cell morphology in chd7 mutants. We further screened for major components of 227
the cartilage matrix. Gene expression analysis at 9 dpf revealed a significant downregulation of 228
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
10
the ECM related gene col2a1a (Fig. 5e). Immunofluorescence showed a striking depletion of 229
Col2a1 in the vertebral cartilage with few cells remaining, which produce a sufficient Col2a1 230
signal (Fig 5f). This pattern was effectively reproduced in all regions of the spine of chd7-/- 231
zebrafish. 232
233
Discussion 234
Efficiency of model 235
CS is a multisystemic disorder likely involving a variety of underlying mechanisms that include 236
CHD7. Therefore, different preclinical models are required to individually identify the various 237
molecular mechanisms. Even though mouse models for CS exist, none have been reported to 238
show the bone phenotype observed in CS patients. Our study presents a chd7 deficient zebrafish 239
that has a bone phenotype resembling that of CS patients, which we used to investigate skeletal 240
development in CS. While scoliosis in CS patients has been attributed to poor muscular 241
development, our study proposes an additional factor involving impaired early mineralization 242
of vertebrae. In line with observations in patients, we observed spinal deformities in chd7-/- 243
zebrafish, indicative of scoliosis.(10-12) Interestingly, our analysis reveals vertebral abnormalities 244
are already present in young larvae of chd7-/- zebrafish. Additional clinical studies are required 245
to investigate the onset of spinal deformities in CS patients. However, early analysis of bone 246
density and spinal development has been suggested by CS guidelines.(45, 46) Detailed analysis 247
may reveal less severe or early onset phenotypes. Similarly, we observed that adult chd7-/- 248
zebrafish, which appeared without spinal deformities on visual inspection, showed signs of 249
spinal abnormalities in microCT analysis, suggesting that mostly mild phenotypes reach this 250
age. This is supported by the high lethality rate shown by Jamadagni et al. (unpublished), yet 251
the underlying cause of lethality is still unclear. 252
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
11
Progressive skeletal disorders such as osteopenia and osteoporosis are characterized by a 253
reduction in bone mineralization over time leading to increased fracture risk.(47) CS patients 254
have been shown to present with significantly reduced bone mineral density.(13) 255
Correspondingly, reduced vertebrae mineralization in chd7-/- zebrafish by 12 dpf was observed 256
in a previous study involving morpholinos in zebrafish which had observed a reduced 257
mineralization of vertebral structures.(27) Confirming previous data, we observed insufficient 258
mineralization of vertebra along the posterior spinal column by 4 weeks of age. In earlier 259
analysis of larvae at 6dpf, we were not able to detect a delay in the onset of mineralization, but 260
already by 9 dpf we saw a significant decrease in the number of calcified vertebrae. 261
Furthermore, mineralization appeared to be less efficient in posterior vertebrae, which could 262
signify an impairment in ossification, since vertebrae in zebrafish undergo ossification from an 263
anterior to posterior direction.(48) An exception from this rule is the Weberian vertebrae C1 and 264
C2, which are ossified later in wildtypes but earlier in chd7-/- mutants. 265
Nutrient dependency 266
Our study showed that the onset of calcification was not delayed in chd7 deficient zebrafish. 267
However, progressing calcification was decreased already 4 days post feeding onset, even 268
though access to food was identical in control and mutants. In contrast, differences in 269
calcification of vertebrae in control and mutants without access to food was not significant. 270
Previous studies have shown that feeding delay until 9 dpf does not affect future fish growth 271
and viability and was therefore our cut-off point.(49) This data suggests that possibly an 272
underlying mechanism involving either reduced uptake or metabolizing nutrients may play a 273
part in the observed phenotype and this link remains to be investigated. Bone calcification and 274
mineralization is dependent on proper nutrition and is often discussed regarding optimal bone 275
health in humans.(50) Insufficient levels of calcium and nutritional uptake are considered an 276
increased risk for bone injury in CS patients.(11) To optimize nutritional uptake, the 277
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
12
supplementation of calcium and other nutrients to counteract poor bone health has been 278
proposed for CS patients.(51) In fact, case reports show supplementation with calcium and 279
vitamin D significantly increased aBMD in CS patients .(52) A possible nutrient deficiency in 280
CS, could be linked to gastrointestinal nutrient uptake difficulties in chd7-/- zebrafish(53), yet 281
this needs to be further investigated. Thus, this possible deficiency in nutrient uptake could 282
further hinder ossification in chd7-/- mutants. We propose using our zebrafish model of CS 283
model to test the effect of guidelines such as calcium and nutrient supplementation, nutrient 284
metabolism and mechanical loading in form of swim training to enhance bone strength in CS. 285
Loss of chd7 reduces osteogenesis in zebrafish 286
Osteoblast precursors commonly express differentiation markers such as runx2 and 287
Osterix/sp7.(54) Upon maturation of these osteoblasts, expression of mineralization related 288
genes such as bglap (osteocalcin) are expressed. Analysis of these three genes showed that they 289
were all significantly downregulated in chd7-/- larvae. This suggests a deficiency in mature 290
osteoblasts, required for the mineralization of the spinal column. Misregulation or loss of theses 291
genes, such as osterix/sp7 has been directly linked to poor bone mineralization in zebrafish and 292
humans.(55-57) Notably, among all osteogenesis related gene expressions significantly reduced 293
in chd7-/- larvae, sost, ctsk, acp5 and col2a1 have been identified as direct regulatory targets of 294
Chd7 by Chip-Seq.(58) Furthermore, the regulation of osteogenesis has been linked to a 295
regulatory complex with PPARγ.(23) Cell culture experiments have proposed the role of CHD7 296
in osteogenesis by inhibiting PPARγ target genes by a protein complex involving CHD7, NLK 297
and SETDB1. Upon loss of CHD7 this complex will fail to inhibit PPARγ target genes and 298
therefore inhibit osteogenesis in preference over adipogenesis.(23) Strikingly, our qPCR results 299
are consistent with downregulation of Pparγ target genes such as runx2a and runx2b. 300
Furthermore, gene expression relevant for bone calcification and remodelling are misregulated 301
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
13
such as sost and ctsk. Thus, our study is the first to indicate the conservation of the proposed 302
pathway shown in cell culture in an in vivo model. 303
Weberian apparatus 304
The Weberian apparatus is a specialized structure that connects the auditory system and the 305
swim bladder in a complex with the first 4 vertebrae to amplify sound vibrations.(33) Our results 306
show that key structures in the apparatus fail to efficiently form and/or mineralize. Two of these 307
structures, the tripus and the parapophysis, are formed to connect to the swim bladder. Both are 308
malformed in all tested chd7-/- mutants to varying degrees. This malformation may imply an 309
inefficiency in its function concerning both the swim bladder and auditory system. Ear 310
abnormalities are a major symptom in CS patients and functionality is affected. Furthermore, 311
the chd7-/- mutant shows swim bladder defects (Breuer et al., unpublished) that could be 312
additionally affecting the morphology in tripus and parapophysis or vice versa. Even though 313
the Weberian apparatus is extremely specialized for carp and carp-like fish (Ostariophysi), it is 314
likely that it underlies the same pathways for osteogenesis as other parts of the spine. 315
Abnormal vertebrae and vBMD variability 316
We examined volumetric bone mineral density (vBMD) and other parameters (e.g. BV, TV) in 317
vertebral bone structures. Most strikingly, we detected large variance in density and all 318
parameters examined, as well as some significant changes in both precaudal and caudal 319
vertebrae. The large variance observed in chd7-/- mutants is in line with CS patients showing 320
varying severity in spinal deformities.(12, 59) Furthermore, previous studies have shown that 321
even minor changes in the described parameters can impact the risk for spinal deformities and 322
fractures in disease and under treatment.(60, 61) 323
Another interesting feature we observed in most chd7-/- mutants, were highly mineralized 324
inclusions, most commonly, localized towards the growth zone of the vertebrae. We did not 325
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
14
identify the source of these highly mineralized structures, but they could be caused by highly 326
localized clusters of osteoblasts during vertebrae mineralization. Another possibility is the 327
occurrence of calcified cartilage, which is more mineralized in comparison to bone. Although 328
speculative, the occurrence of mineralized cartilage could be due to incomplete endochondral 329
ossification, which would be in line with the reduction in sp7/osterix and col2a1 expression(62, 330
63) Since these inclusions are highly mineralized, it seems likely that their appearance impairs 331
the bone structurally as higher mineralized collagen is stiffer, yet less tough.(64) It will be of 332
interest to see if future studies can take a closer look at these previously unidentified features. 333
While these minor defects could be involved in the structural integrity of the spine in CS 334
patients, these minor features may be missed during dual-energy x-ray absorptiometry (DEXA) 335
screens of CS patients when assessing aBMD. 336
Analysis revealed a decreased mineralization in the caudal vertebral arches of the mutant fish. 337
This likely reduces the mechanical competence of these vertebrae in comparison to healthy 338
vertebrae. A compromised mechanical competence alongside low BMD also increases fracture 339
risk and can be an underlying cause of scoliosis.(65-67) Unfortunately, natural history BMD data 340
of CS patients throughout early development is not available, to the best of our knowledge. 341
However, aBMD analysis of patients with idiopathic scoliosis in CS via DEXA scans has been 342
studied and shows significant reduction in some cases.(13, 68) In general, there is an increased 343
fracture risk and incidence of scoliosis corresponding to decreased BMD, such as in osteopenia 344
and osteoporosis.(47, 66) In line with this reduced mineralization an increased fracture and injury 345
risk has been reported for CS patients.(13) Accordingly, proper observation and focus on early 346
DEXA analysis is considered for CS guidelines.(45, 46) 347
Our analysis throughout development of chd7-/- deficient zebrafish suggests that only mild 348
phenotypes survive onto later stages with severity decreasing as age progresses. However, the 349
high lethality in both patients and our model is connected to a multisystemic phenotype in which 350
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
15
one mechanism is unlikely to be exclusive underlying cause for this characteristic. The link 351
between Chd7 and severe spinal and vertebrae abnormalities, including kyphosis, fused 352
vertebrae and hemivertebrae has been observed in our samples.(7, 9) 353
Collagen and chondrocyte depletion as additional risk factor 354
Our data in chd7 deficient zebrafish indicate that reduced vBMD may not be the only 355
underlying risk factor for idiopathic scoliosis in CS patients. Spinal integrity may be weakened 356
by a decrease in chondrocyte marker col2a1 expression in the vertebral cartilage of chd7 357
deficient zebrafish. Col2a1 is required in the ECM of various cell types including osteoblasts, 358
chondrocytes, external ligament connective tissue cells, as well as notochord basal cells and 359
expressed in the corresponding tissues.(69, 70) Each cell type is required for proper integrity and 360
could likely contribute to spinal injuries and malformations. Furthermore, col2a1a expression 361
is observed throughout ossification of the notochord.(71) Reduction in this collagen matrix in 362
the notochord can affect proper calcification. Col2a1a is expressed in osteoblasts and 363
chondrocytes of teleosts and regulated by sox9, which in turn is known to be regulated by 364
Chd7.(72-74) The hypertrophic morphology of the chondrocytes seen in our study may indicate 365
signs of a weakened matrix, as well as altered matrix composition and may contribute to an 366
increased risk in osteoporosis.(75, 76) How this regulatory defect may influence other 367
cartilage/skeletal regions such as the craniofacial cartilage remains to be investigated. 368
In summary, our study is the first to identify skeletal abnormalities linked to chd7 deficiency. 369
These abnormalities include morphological changes in vertebrae, reduced bone mineralization, 370
impaired osteoblast activity, highly mineralized inclusions and Col2a1 deficiency in the chd7-371
/- zebrafish model organism. The use of this preclinical model will allow for future studies to 372
further understand the underlying mechanisms of spinal deformities in CS. 373
374
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
16
Materials and Methods 375
Zebrafish husbandry 376
Wildtype (Cntrl) and mutant (chd7-/-) zebrafish were kept at 28°C in a 12h/12h dark/light cycle 377
and maintained in accordance with Westerfield et al.(77). All zebrafish in this study were fed a 378
steady diet of Skretting®GemmaMicro starting at 5dpf. Embryos were raised at 28.5°C and 379
staged as previously described by Kimmel et al.(78) All experiments were performed in with the 380
guidelines of the Canadian Council for Animal Care and the local ethics committee. 381
Skeletal stainings and analysis of larvae 382
Calcein staining was done as described(38). In short, zebrafish larvae at 9dpf were exposed to 383
calcein (2g/L) (Sigma-Aldrich) in water for 10 minutes. Larvae were then washed at least 3 384
times in fish water for 10 minutes to remove excess calcein. Larvae were then anesthetized 385
using tricaine before being imaged using an AxioZoom V16 (Zeiss). For nutrition dependency 386
tests, larvae were either deprived of food until 9 dpf or fed a regular Skretting®GemmaMicro 387
diet starting at 5 dpf. 388
Alizarin Red was performed as previously shown(79). In brief, larvae were fixed for 2 hours with 389
2% PFA. The PFA was then removed by washes in PBS 3x10 minutes and then placed in 25% 390
glycerol/0.1%KOH. Finally, larvae were stained with 0,05% Alizarin Red (Sigma-Aldrich) in 391
H2O for 30 minutes and stored in 50% glycerol in 0.1%KOH. 392
MicroCT Imaging 393
Adult zebrafish were euthanized using tricaine (MS-222, Sigma-Aldrich). Tissue was fixed in 394
4% PFA prior to microCT analysis. Ex vivo microCT at an isotropic voxel size of 10.6 µm 395
(SkyScan1276, Bruker, Kontich, Belgium) was performed to assess differences in bone mass, 396
mineral density, and microstructure. The following scanning parameters were used: peak 397
voltage of 55 kVp, Aluminum 0.25 mm filter, source current of 200 µA, 0.3° rotational steps 398
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
17
for full 360° with no frame averaging. Images were reconstructed using standard reconstruction 399
algorithms provided with the microCT. Spinal deformities of controls and mutants were 400
analyzed of 1-year (n=8 for Cntrl and n=8 chd7-/-) and 2-year-old (n=5 for Cntrl and n=5 chd7-401
/-) zebrafish. 402
Analysis and bone density/volume/angles 403
From the reconstructed microCT images of the whole spine, the Weberian apparatus, as well as 404
vertebrae 7 (abdominal) and 20 (caudal) were segmented. Further, the intercalarium, the tripus 405
as well as the parapophysis were segmented from the Weberian apparatus. For the vertebrae, 406
the neural as well as the hemal arch were segmented from the vertebral body. To differentiate 407
between bone and background, a density-based threshold of 0.41 gHA/cm3 determined by 408
Otsu’s method(80) was used. For each segmented bone in the Weberian apparatus outcomes 409
included: bone volume BV (mm3), tissue volume TV (mm3), bone volume fraction BV/TV 410
(mm3/mm3) and volumetric bone mineral density vBMD (gHA/cm3). For abdominal and caudal 411
vertebrae the outcomes included: bone volume BV, tissue volume TV, bone volume fraction 412
BV/TV, total bone mineral density T.vBMD, vertebral body vBMD VB.BMD, vertebral arch 413
vBMD VA.BMD, as well as arch-body angle, body angle, arch opening angle and rising angles 414
(see Fig.2a and 3a). For caudal vertebrae, the angle outcomes were calculated for hemal as well 415
as neural arches. 416
qRT-PCR 417
Total RNA was isolated from 9 dpf old zebrafish larvae using TriReagent. 1µg of RNA was 418
used for cDNA synthesis using cDNA vilo kit. qRT-PCR was performed with SYBR Green 419
mix (BIORAD) with a Lightcycler96® (Roche). Gene expression was analyzed relative to the 420
housekeeping gene elf1α. Primers used for runx2a, runx2b, sost, bglap, sp7/osterix, mgp, 421
postna, acp5a, ctsk, col2a1 and mgp are shown in supplementals Table 5. 422
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
18
Spinal sections/Stainings/Immunohistochemistry 423
The zebrafish were euthanized in tricaine (MS-222, Sigma-Aldrich) and fixed for 3 days in 4% 424
paraformaldehyde at 4°C, after the abdomen was opened to ensure proper fixation. The spinal 425
cords were dissected and decalcified with EDTA 10 % for 3 days under agitation at RT. After 426
fixation and decalcification, the zebrafish were embedded in paraffin. Longitudinal sections 427
(5μm) were obtained and were deparaffinized in xylene and were rehydrated in a graded series 428
of ethanol. The slides were stained with hematoxylin (STATLAB Medical Products, LLC) for 429
4 min, and washed with alcohol-acid, and were rinsed with tap water. In the blueing step, the 430
slides were soaked in saturated lithium carbonate solution for 10 sec, and then rinsed with tap 431
water. Finally, staining was performed with eosin Y (STATLAB Medical Products, LLC) for 2 432
min and mounted with Permount™ mounting medium. 433
For Safranin O/Fast green, after dehydration, the slides were stained with Weigert’s 434
hematoxylin (Sigma-Aldrich) for 4 min, wash under tap water, immersed in alcohol–acid for 5 435
sec, then washed under water. The slides were stained with Fast green 0.02 % for 2 min, washed 436
with acetic water 1% for 20 sec, then stained directly in safranin 0.01 % for 5 min and mounted 437
with Permount™ mounting medium. Images were taken using an AxioZoom V16 (Zeiss) and 438
area and number of nuclei were determined using ImageJ. 439
Immunofluorescence for Col2a1 was performed as previously shown69. In short, Epitope 440
demasking was performed by digestion with Proteinase K (20µg/ml). Tissue was blocked with 441
NGS and samples treated with primary antibody for Col2a1 (1 in 10; Developmental studies 442
hybridoma bank) and secondary antibody Goat-anti-Mouse Alexafluor 488 (1 in 300). Counter 443
stain was done using DAPI-mounting medium. Images were taken using a LSM780 (Zeiss). 444
Statistics 445
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
19
Statistical analysis was carried out using Prism-GraphPad® (GraphPad, San Diego, CA, USA). 446
qPCR data and nuclei quantification were tested with student’s t-test. Calcification analysis in 447
larvae was performed with one-way ANOVA and post-hoc Tukey Test. MiroCT data of Control 448
vs chd7-/- mutants was analyzed using student’s t-test with Welch’s correction and Variance 449
determined by F-test analysis. Significance was determined at *<0,05; **<0,01 and ***<0,001 450
Acknowledgements 451
The authors would like to thank Beatrice Steyn for experimental support and Mark Lepik for 452
graphic design. SP is supported by the Natural Science and Engineering Research Council 453
(NSERC), Canadian Institutes of Health Research (CIHR), an ALS Canada-Brain Canada 454
Career Transition Award and a FRQS Junior 1 research scholar. BMW is supported by FRQS 455
Programme de bourses de chercheur. MR is supported by FRQS Programme de bourses de 456
chercheur. We also thank Shriners Hospital for Children and the CIHR (165939) for financial 457
support. 458
Author roles 459
MB, MR, BMW and SP designed the study. MB, MR and CZ performed the experiments. MB 460
and MR drafted the manuscript. MB, MR, CZ, BMW, SP edited the mansucript. 461
462
References 463
1. Wise CA, Gao X, Shoemaker S, Gordon D, Herring JA. Understanding genetic factors 464 in idiopathic scoliosis, a complex disease of childhood. Curr Genomics. 2008;9(1):51-9. 465
2. Grauers A, Einarsdottir E, Gerdhem P. Genetics and pathogenesis of idiopathic 466 scoliosis. Scoliosis Spinal Disord. 2016;11:45. 467 3. Hall BD. Choanal atresia and associated multiple anomalies. J Pediatr. 1979;95(3):395-468 8. 469 4. Hittner HM, Hirsch NJ, Kreh GM, Rudolph AJ. Colobomatous microphthalmia, heart 470
disease, hearing loss, and mental retardation--a syndrome. J Pediatr Ophthalmol Strabismus. 471 1979;16(2):122-8. 472
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
20
5. Pagon RA, Graham JM, Jr., Zonana J, Yong SL. Coloboma, congenital heart disease, 473
and choanal atresia with multiple anomalies: CHARGE association. J Pediatr. 1981;99(2):223-474 7. 475
6. Blake KD, Salem-Hartshorne N, Daoud MA, Gradstein J. Adolescent and adult issues 476 in CHARGE syndrome. Clin Pediatr (Phila). 2005;44(2):151-9. 477 7. Jongmans MC, Admiraal RJ, van der Donk KP, Vissers LE, Baas AF, Kapusta L, et al. 478 CHARGE syndrome: the phenotypic spectrum of mutations in the CHD7 gene. J Med Genet. 479 2006;43(4):306-14. 480
8. Verloes A. Updated diagnostic criteria for CHARGE syndrome: a proposal. Am J Med 481 Genet A. 2005;133A(3):306-8. 482 9. Sanlaville D, Verloes A. CHARGE syndrome: an update. Eur J Hum Genet. 483 2007;15(4):389-99. 484 10. Doyle C, Blake K. Scoliosis in CHARGE: a prospective survey and two case reports. 485
Am J Med Genet A. 2005;133A(3):340-3. 486
11. Forward KE, Cummings EA, Blake KD. Risk factors for poor bone health in adolescents 487
and adults with CHARGE syndrome. Am J Med Genet A. 2007;143A(8):839-45. 488 12. Gao X, Gordon D, Zhang D, Browne R, Helms C, Gillum J, et al. CHD7 gene 489 polymorphisms are associated with susceptibility to idiopathic scoliosis. Am J Hum Genet. 490 2007;80(5):957-65. 491
13. Kumaran A, Kirk J. Endocrine manifestations of CHARGE syndrome. Endocrine 492 Abstracts. 2013;33:P80. 493
14. Lalani SR, Safiullah AM, Fernbach SD, Harutyunyan KG, Thaller C, Peterson LE, et 494 al. Spectrum of CHD7 mutations in 110 individuals with CHARGE syndrome and genotype-495 phenotype correlation. Am J Hum Genet. 2006;78(2):303-14. 496
15. Mitchell JA, Giangiacomo J, Hefner MA, Thelin JW, Pickens JM. Dominant CHARGE 497 association. Ophthalmic Paediatr Genet. 1985;6(1-2):271-6. 498
16. Delahaye A, Sznajer Y, Lyonnet S, Elmaleh-Berges M, Delpierre I, Audollent S, et al. 499
Familial CHARGE syndrome because of CHD7 mutation: clinical intra- and interfamilial 500
variability. Clin Genet. 2007;72(2):112-21. 501 17. Jones KM, Saric N, Russell JP, Andoniadou CL, Scambler PJ, Basson MA. CHD7 502 maintains neural stem cell quiescence and prevents premature stem cell depletion in the adult 503
hippocampus. Stem Cells. 2015;33(1):196-210. 504
18. Ohta S, Yaguchi T, Okuno H, Chneiweiss H, Kawakami Y, Okano H. CHD7 promotes 505 proliferation of neural stem cells mediated by MIF. Mol Brain. 2016;9(1):96. 506 19. Okuno H, Renault Mihara F, Ohta S, Fukuda K, Kurosawa K, Akamatsu W, et al. 507 CHARGE syndrome modeling using patient-iPSCs reveals defective migration of neural crest 508 cells harboring CHD7 mutations. Elife. 2017;6. 509
20. Writzl K, Cale CM, Pierce CM, Wilson LC, Hennekam RC. Immunological 510 abnormalities in CHARGE syndrome. Eur J Med Genet. 2007;50(5):338-45. 511 21. Feng W, Shao C, Liu HK. Versatile Roles of the Chromatin Remodeler CHD7 during 512
Brain Development and Disease. Front Mol Neurosci. 2017;10:309. 513 22. Janssen N, Bergman JE, Swertz MA, Tranebjaerg L, Lodahl M, Schoots J, et al. 514 Mutation update on the CHD7 gene involved in CHARGE syndrome. Hum Mutat. 515 2012;33(8):1149-60. 516
23. Takada I, Kouzmenko AP, Kato S. Wnt and PPARgamma signaling in 517 osteoblastogenesis and adipogenesis. Nat Rev Rheumatol. 2009;5(8):442-7. 518 24. Chen Y, Wang M, Chen D, Wang J, Kang N. Chromatin remodeling enzyme CHD7 is 519 necessary for osteogenesis of human mesenchymal stem cells. Biochem Biophys Res Commun. 520 2016;478(4):1588-93. 521
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
21
25. Ogier JM, Carpinelli MR, Arhatari BD, Symons RC, Kile BT, Burt RA. CHD7 522
deficiency in "Looper", a new mouse model of CHARGE syndrome, results in ossicle 523 malformation, otosclerosis and hearing impairment. PLoS One. 2014;9(5):e97559. 524
26. Hurd EA, Capers PL, Blauwkamp MN, Adams ME, Raphael Y, Poucher HK, et al. Loss 525 of Chd7 function in gene-trapped reporter mice is embryonic lethal and associated with severe 526 defects in multiple developing tissues. Mamm Genome. 2007;18(2):94-104. 527 27. Patten SA, Jacobs-McDaniels NL, Zaouter C, Drapeau P, Albertson RC, Moldovan F. 528 Role of Chd7 in zebrafish: a model for CHARGE syndrome. PLoS One. 2012;7(2):e31650. 529
28. Busse B, Galloway JL, Gray RS, Harris MP, Kwon RY. Zebrafish: An Emerging Model 530 for Orthopedic Research. J Orthop Res. 2020;38(5):925-36. 531 29. Carnovali M, Banfi G, Mariotti M. Zebrafish Models of Human Skeletal Disorders: 532 Embryo and Adult Swimming Together. Biomed Res Int. 2019;2019:1253710. 533 30. Ofer L, Dean MN, Zaslansky P, Kult S, Shwartz Y, Zaretsky J, et al. A novel 534
nonosteocytic regulatory mechanism of bone modeling. PLoS Biol. 2019;17(2):e3000140. 535
31. Hayes AJ, Reynolds S, Nowell MA, Meakin LB, Habicher J, Ledin J, et al. Spinal 536
deformity in aged zebrafish is accompanied by degenerative changes to their vertebrae that 537 resemble osteoarthritis. PLoS One. 2013;8(9):e75787. 538 32. Bergen DJM, Kague E, Hammond CL. Zebrafish as an Emerging Model for 539 Osteoporosis: A Primary Testing Platform for Screening New Osteo-Active Compounds. Front 540
Endocrinol (Lausanne). 2019;10:6. 541 33. Dahdul WM, Lundberg JG, Midford PE, Balhoff JP, Lapp H, Vision TJ, et al. The 542
teleost anatomy ontology: anatomical representation for the genomics age. Syst Biol. 543 2010;59(4):369-83. 544 34. Suniaga S, Rolvien T, Vom Scheidt A, Fiedler IAK, Bale HA, Huysseune A, et al. 545
Increased mechanical loading through controlled swimming exercise induces bone formation 546 and mineralization in adult zebrafish. Sci Rep. 2018;8(1):3646. 547
35. Haga Y, Dominique VJ, 3rd, Du SJ. Analyzing notochord segmentation and 548
intervertebral disc formation using the twhh:gfp transgenic zebrafish model. Transgenic Res. 549
2009;18(5):669-83. 550 36. Morin-Kensicki EM, Melancon E, Eisen JS. Segmental relationship between somites 551 and vertebral column in zebrafish. Development. 2002;129(16):3851-60. 552
37. Spoorendonk KM, Hammond CL, Huitema LFA, Vanoevelen J, Schulte-Merker S. 553
Zebrafish as a unique model system in bone research: the power of genetics and in vivo imaging. 554 2010;26(2):219-24. 555 38. Xi Y, Chen D, Sun L, Li Y, Li L. Characterization of zebrafish mutants with defects in 556 bone calcification during development. Biochem Biophys Res Commun. 2013;440(1):132-6. 557 39. MacRae CA, Peterson RT. Zebrafish as tools for drug discovery. Nature reviews Drug 558
discovery. 2015;14(10):721-31. 559 40. Fleming A, Sato M, Goldsmith P. High-throughput in vivo screening for bone anabolic 560 compounds with zebrafish. J Biomol Screen. 2005;10(8):823-31. 561
41. Boswell CW, Ciruna B. Understanding Idiopathic Scoliosis: A New Zebrafish School 562 of Thought. Trends Genet. 2017;33(3):183-96. 563 42. Guo L, Ikegawa S, Shukunami C. Emergence of Zebrafish as a Model System for 564 Understanding Human Scoliosis. In: Hirata H, Iida A, editors. Zebrafish, Medaka, and Other 565
Small Fishes: New Model Animals in Biology, Medicine, and Beyond. Singapore: Springer 566 Singapore; 2018. p. 217-34. 567 43. Gray RS, Wilm TP, Smith J, Bagnat M, Dale RM, Topczewski J, et al. Loss of col8a1a 568 function during zebrafish embryogenesis results in congenital vertebral malformations. Dev 569 Biol. 2014;386(1):72-85. 570
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
22
44. Gistelinck C, Kwon RY, Malfait F, Symoens S, Harris MP, Henke K, et al. Zebrafish 571
type I collagen mutants faithfully recapitulate human type I collagenopathies. Proc Natl Acad 572 Sci U S A. 2018;115(34):E8037-E46. 573
45. de Geus CM, Free RH, Verbist BM, Sival DA, Blake KD, Meiners LC, et al. Guidelines 574 in CHARGE syndrome and the missing link: Cranial imaging. Am J Med Genet C Semin Med 575 Genet. 2017;175(4):450-64. 576 46. Trider CL, Arra-Robar A, van Ravenswaaij-Arts C, Blake K. Developing a CHARGE 577 syndrome checklist: Health supervision across the lifespan (from head to toe). Am J Med Genet 578
A. 2017;173(3):684-91. 579 47. Aspray TJ, Hill TR. Osteoporosis and the Ageing Skeleton. Subcell Biochem. 580 2019;91:453-76. 581 48. Bensimon-Brito A, Cardeira J, Cancela ML, Huysseune A, Witten PE. Distinct patterns 582 of notochord mineralization in zebrafish coincide with the localization of Osteocalcin isoform 583
1 during early vertebral centra formation. BMC Dev Biol. 2012;12:28. 584
49. Hernandez RE, Galitan L, Cameron J, Goodwin N, Ramakrishnan L. Delay of Initial 585
Feeding of Zebrafish Larvae Until 8 Days Postfertilization Has No Impact on Survival or 586 Growth Through the Juvenile Stage. Zebrafish. 2018;15(5):515-8. 587 50. O'Keefe JH, Bergman N, Carrera-Bastos P, Fontes-Villalba M, DiNicolantonio JJ, 588 Cordain L. Nutritional strategies for skeletal and cardiovascular health: hard bones, soft arteries, 589
rather than vice versa. Open Heart. 2016;3(1):e000325. 590 51. Blake KD, Hudson AS. Gastrointestinal and feeding difficulties in CHARGE syndrome: 591
A review from head-to-toe. Am J Med Genet C Semin Med Genet. 2017;175(4):496-506. 592 52. Foppiani L, Maffe A, Forzano F. CHARGE syndrome as unusual cause of 593 hypogonadism: endocrine and molecular evaluation. Andrologia. 2010;42(5):326-30. 594
53. Cloney K, Steele SL, Stoyek MR, Croll RP, Smith FM, Prykhozhij SV, et al. Etiology 595 and functional validation of gastrointestinal motility dysfunction in a zebrafish model of 596
CHARGE syndrome. FEBS J. 2018;285(11):2125-40. 597
54. Ortuno MJ, Susperregui AR, Artigas N, Rosa JL, Ventura F. Osterix induces Col1a1 598
gene expression through binding to Sp1 sites in the bone enhancer and proximal promoter 599 regions. Bone. 2013;52(2):548-56. 600 55. Kague E, Roy P, Asselin G, Hu G, Simonet J, Stanley A, et al. Osterix/Sp7 limits cranial 601
bone initiation sites and is required for formation of sutures. Dev Biol. 2016;413(2):160-72. 602
56. Lapunzina P, Aglan M, Temtamy S, Caparros-Martin JA, Valencia M, Leton R, et al. 603 Identification of a frameshift mutation in Osterix in a patient with recessive osteogenesis 604 imperfecta. Am J Hum Genet. 2010;87(1):110-4. 605 57. Niu P, Zhong Z, Wang M, Huang G, Xu S, Hou Y, et al. Zinc finger transcription factor 606 Sp7/Osterix acts on bone formation and regulates col10a1a expression in zebrafish. Science 607
Bulletin. 2017;62(3):174-84. 608 58. Rouillard AD, Gundersen GW, Fernandez NF, Wang Z, Monteiro CD, McDermott MG, 609 et al. The harmonizome: a collection of processed datasets gathered to serve and mine 610
knowledge about genes and proteins. Database (Oxford). 2016;2016. 611 59. Tilley MK, Justice CM, Swindle K, Marosy B, Wilson AF, Miller NH. CHD7 gene 612 polymorphisms and familial idiopathic scoliosis. Spine (Phila Pa 1976). 2013;38(22):E1432-6. 613 60. Faulkner KG. Bone matters: are density increases necessary to reduce fracture risk? J 614
Bone Miner Res. 2000;15(2):183-7. 615 61. Li XF, Li H, Liu ZD, Dai LY. Low bone mineral status in adolescent idiopathic 616 scoliosis. Eur Spine J. 2008;17(11):1431-40. 617 62. Oh JH, Park SY, de Crombrugghe B, Kim JE. Chondrocyte-specific ablation of Osterix 618 leads to impaired endochondral ossification. Biochem Biophys Res Commun. 619 2012;418(4):634-40. 620
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
23
63. Weigele J, Franz-Odendaal TA. Functional bone histology of zebrafish reveals two 621
types of endochondral ossification, different types of osteoblast clusters and a new bone type. 622 J Anat. 2016;229(1):92-103. 623
64. Currey JD. Bones: Structure and Mechanics: Princeton University Press; 2006. 624 65. Zimmermann EA, Busse B, Ritchie RO. The fracture mechanics of human bone: 625 influence of disease and treatment. Bonekey Rep. 2015;4:743. 626 66. Sarioglu O, Gezer S, Sarioglu FC, Koremezli N, Kara T, Akcali O, et al. Evaluation of 627 vertebral bone mineral density in scoliosis by using quantitative computed tomography. Pol J 628
Radiol. 2019;84:e131-e5. 629 67. Wang Z, Chen H, Yu YE, Zhang J, Cheuk KY, Ng BK, et al. Unique local bone tissue 630 characteristics in iliac crest bone biopsy from adolescent idiopathic scoliosis with severe spinal 631 deformity. Sci Rep. 2017;7:40265. 632 68. Modi A, Piziak V. MON-355 Charge Syndrome: Unusual Cause of Hypogonadism 633
Leading to Osteoporosis. J Endocr Soc. 2020;4(Suppl 1):MON-355. 634
69. Kaneko T, Freeha K, Wu X, Mogi M, Uji S, Yokoi H, et al. Role of notochord cells and 635
sclerotome-derived cells in vertebral column development in fugu, Takifugu rubripes: 636 histological and gene expression analyses. Cell Tissue Res. 2016;366(1):37-49. 637 70. Jian QL, HuangFu WC, Lee YH, Liu IH. Age, but not short-term intensive swimming, 638 affects chondrocyte turnover in zebrafish vertebral cartilage. PeerJ. 2018;6:e5739. 639
71. Dale RM, Topczewski J. Identification of an evolutionarily conserved regulatory 640 element of the zebrafish col2a1a gene. Dev Biol. 2011;357(2):518-31. 641 72. Akiyama H, Chaboissier MC, Martin JF, Schedl A, de Crombrugghe B. The transcription factor 642 Sox9 has essential roles in successive steps of the chondrocyte differentiation pathway and is required 643 for expression of Sox5 and Sox6. Genes Dev. 2002;16(21):2813-28. 644 73. Bajpai R, Chen DA, Rada-Iglesias A, Zhang J, Xiong Y, Helms J, et al. CHD7 cooperates with PBAF 645 to control multipotent neural crest formation. Nature. 2010;463(7283):958-62. 646 74. Eames BF, Amores A, Yan YL, Postlethwait JH. Evolution of the osteoblast: skeletogenesis in gar 647 and zebrafish. BMC Evol Biol. 2012;12:27. 648 75. Hall AC. The Role of Chondrocyte Morphology and Volume in Controlling Phenotype-649 Implications for Osteoarthritis, Cartilage Repair, and Cartilage Engineering. Curr Rheumatol Rep. 650 2019;21(8):38. 651 76. Lian C, Wang X, Qiu X, Wu Z, Gao B, Liu L, et al. Collagen type II suppresses articular chondrocyte 652 hypertrophy and osteoarthritis progression by promoting integrin beta1-SMAD1 interaction. Bone Res. 653 2019;7:8. 654 77. Westerfield M. The zebrafish book : a guide for the laboratory use of zebrafish (Danio rerio) 655 4th ed. [Text.]. Eugene, OR: Institute of Neuroscience, University of Oregon ,; 2000 [Available from: 656 http://zfish.uoregon.edu/zf_info/zfbook/zfbk.html. 657 78. Kimmel CB, Ballard WW, Kimmel SR, Ullmann B, Schilling TF. Stages of embryonic development 658 of the zebrafish. Dev Dyn. 1995;203(3):253-310. 659 79. Aceto J, Nourizadeh-Lillabadi R, Maree R, Dardenne N, Jeanray N, Wehenkel L, et al. Zebrafish 660 Bone and General Physiology Are Differently Affected by Hormones or Changes in Gravity. PLoS One. 661 2015;10(6):e0126928. 662 80. Otsu N. A Threshold Selection Method from Gray-Level Histograms. IEEE Transactions on 663 Systems, Man, and Cybernetics. 1979;9(1):62-6. 664
665
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
24
666
Fig 1. Mineralization in larval and juvenile chd7-/- mutants 667
a. Dorsal view of control and chd7-/- larvae at 9dpf showing severe spinal deformities in 43 out of 110 668
larvae screened. Deformities were present at both precaudal and caudal region (arrows). b. Lateral view 669
of calcein staining at 9dpf showing mineralization of vertebrae with access to food starting at 5dpf and 670
c. calcein staining at 9dpf during starvation. Arrow indicates mineralization at the first 4 (Weberian) 671
vertebrae d. Number of vertebrae in wildtype and chd7-/- (fed and starved) larvae showing at least partial 672
calcification signal as tested by calcein and e. Number of weberian vertebrae (vertebrae 1-4) showing 673
complete calcification (see arrows in b) f. Lateral view of 4 week old zebrafish stained with alizarin red 674
to screen mineralized vertebrae, showing decreased calcification towards posterior region. Bottom right 675
corners show magnification of selected caudal vertebrae. g. RT-qPCR of osteogenesis related genes 676
screened at 9dpf, showing relative fold change. (Significance: *p<0.05; **p<0.01) 677
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
25
678
Fig 2. Weberian apparatus 679
a. Representative microCT overview image of the spinal cord from a 1-year old control and a chd7-/- 680
adult zebrafish b. MicroCT Image of the Weberian apparatus from a 1-year old control and a chd7-/- 681
mutants, indicating structures of intercalarium (red), tripus (yellow) and parapophysis (blue). c-h. 682
vBMD and BV/TV of intercalarium (c, d) tripus (e, f) and parapophysis (g, h) (n=8/genotype were 683
analyzed). Significance of student t-test with Welch’s correction is included in the graphs with: *p<0.05; 684
**p<0.01. F-test results are indicated underneath as Fdegree of freedom numerator, degree of freedom denominator (F-value); 685
p-value: *p<0.05; **p<0.01; ***p<0.001 686
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
26
687
Fig 3. Precaudal vertebrae show inefficiency for proper mineralization 688
a. Sketch diagram of precaudal vertebrae (frontal and lateral) indicating structure and measured angles. 689
b-c. (top) Individual rendering of precaudal vertebrae, frontal view of tissue over 0.41 g HA/cm3 690
threshold and (bottom) vBMD intensity map (Range blue-red; 0.41 – 1.00 g HA/cm3) showing 691
morphological abnormalities and growth zone malformations with highly mineralized inclusions 692
(arrow) d. vBMD of whole vertebrae showing increasing density with increasing age and e. increasing 693
ratio of vBMD arch/vertebrae body. f-h neural arch angle, BV, TV and ratio BV/TV. (n=8/genotype 694
were analyzed). Significance of student t-test with Welch’s correction is included in the graphs with: 695
*p<0.05; **p<0.01. F-test results are indicated underneath as Fdegree of freedom numerator, degree of freedom denominator 696
(F-value); p-value: *p<0.05; **p<0.01; ***p<0.001 697
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
27
698
Fig 4. Caudal vertebrae show abnormal structure 699
a. Sketch diagram of precaudal vertebrae indicating structure and measured angles (frontal and lateral). 700
b-c. (top) Individual rendering of caudal vertebrae, frontal view of tissue over 0.41 g HA/cm3 threshold 701
and (bottom) vBMD intensity map (Range blue-red; 0.41 – 1.00 g HA/cm3) showing morphological 702
abnormalities and growth zone malformations with highly mineralized inclusions (arrow) d. vBMD of 703
whole vertebrae and E. increasing ratio of vBMD arch/vertebrae body. f-h neural arch angle, BV, TV 704
and ratio BV/TV. (n=8/genotype were analyzed). Significance of student t-test with Welch’s correction 705
is included in the graphs with: *p<0.05; **p<0.01. F-test results are indicated underneath as Fdegree of 706
freedom numerator, degree of freedom denominator (F-value); p-value: *p<0.05; **p<0.01; ***p<0.001 707
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint
28
708
Fig 5. ECM Collagen deficiency in 1-year-old chd7-/- adults 709
a. H&E staining of precaudal vertebrae section of Cntrl and chd7-/- mutants b. Total nuclei count in 710
precaudal vertebral cartilage detected in H&E staining per 100μm2 c. Safranin O/fast green staining of 711
precaudal vertebrae section Cntrl and chd7-/- mutants showing cartilage in red. d. Chondrocyte nuclei 712
count in precaudal vertebral cartilage detected in Safranin O/fast green staining per 100μm2 e. qPCR of 713
col2a1a at 9dpf. f. Immunofluorescence of Col2a1 in precaudal vertebrae sections of Cntrl and chd7-/- 714
mutants with Col2a1 in green and DAPI in blue. Scale bar presents 100μm. Significance of student t-715
test with: *p<0.05; **p<0.01 ***p<0.001 716
(which was not certified by peer review) is the author/funder. All rights reserved. No reuse allowed without permission. The copyright holder for this preprintthis version posted July 11, 2020. . https://doi.org/10.1101/2020.07.10.197533doi: bioRxiv preprint